Bob Murphy and Karl Smith are quickly becoming my favorite bloggers to follow. This time, Bob starts with a thought experiment and Karl provides a good response. Good thinking without any cheap shots on the opposition... a rare thing.
So the gnome attack thought experiment has to do with the capital heterogeneity issue.
I'm not sure I entirely get Bob's point here. If gnomes attack like this, how would Keynesians talk about it? They'd say it's a sharp fall in the marginal efficiency of capital, wouldn't they? Bob acts like Keynesians have no way of talking about specific capital that doesn't work well, so they try to fit the square aggregate demand peg into the round hole in the observed data. That seems wrong to me.
Now - as I've pointed out previously - Keynesianism relies on heterogeneous capital*. So does the Austrian school. Does heterogeneity of capital play a different role in each? Sure - each uses capital heterogeneity differently, and I've long said that the two views of capital heterogeneity could be profitably synthesized.
Austrians basically say "capital is specific - distortions of price signals lead to unsustainable capital investments which will have to be worked out down the road" (the prime distortion of interest being, of course, the distortion of the time structure).
Keynesians basically say "capital is specific - saving doesn't guarantee investment because (1.) non-investment savings vehicles are available, and (2.) savers are not obligated to place specific orders for future demand, which are required because capital is specific and investors won't invest just because 'savers will consume something at some time in the future'."
Austrians often (always?) wave their hands at the implications of liquidity preference for equilibria below full employment. Keynesians wave their hands at specific distortions of the capital structure, but they at least have a way of incorporating it into their models - namely the marginal efficiency of capital. There are lots of ways that the capital structure could be distorted. Gnome attacks. Zombie attacks (not as effective as gnomes). Ninja attacks (this would probably be even more effective than gnomes). Pirate attacks, etc. Of course if you really stretch your imagination you could also consider technological advance and obsolescence, political instability, interest rate distortion of the time structure, changes in regulations, wars, bottlenecks in energy supplies, and other problems messing up a heterogeneous capital structure too. It's not clear to me why we would want to stake a business cycle theory on any single problem with the capital structure, particularly when we typically think entrepreneurs are more or less rational and profit maximizing. I prefer the more general Keynesian approach which explains how - given a particular marginal efficiency of capital - the economy can sit at an equilibrium below full employment.
This hasn't prevented Keynesians from identifying specific problems when they're relevant. When the oil embargo happened, didn't everybody agree there was an oil embargo? When the dustbowl happened, didn't we all agree about the role of the dustbowl? Haven't we all agreed about the role that mortgage backed securities played in distorting the capital structure more recently? I don't see why Bob thinks this would be such a mystery. If gnomes struck I think we'd all agree about the gnomes too.
But if you were to ask me how - given the attack of the gnomes - the economy would respond to a new marginal efficiency of capital, I still think a Keynesian model is going to give you the best answer. You can't blame a model for not anticipating every potential exogenous monkey wrench.
*Even if the theory didn't rely on it, everyone thinks capital is heterogeneous so this blustering about Austrians being unique in having this insight is a moot point.
The worst of both worlds...
UPDATE: I've been told we need alien gnomes too: